Langue: en

Version: 66003 (mandriva - 22/10/07)

Section: 8 (Commandes administrateur)

BSD mandoc


dccproc - Distributed Checksum Clearinghouse Procmail Interface


dccproc -words [-VdAQCHER ] [-h homedir ] [-m map ] [-w whiteclnt ] [-T tmpdir ] [-a IP-address ] [-f env_from ] [-t targets ] [-x exitcode ]
[-c type, [log-thold, ] rej-thold ] [-g [not- ] type ] [-S header ]
[-i infile ] [-o outfile ] [-l logdir ] [-B dnsbl-option ] [-L ltype,facility.level ]


Dccproc copies a complete SMTP message from standard input or a file to standard output or another file. As it copies the message, it computes the DCC checksums for the message, reports them to a DCC server, and adds a header line to the message. Another program such as procmail(1) can use the added header line to filter mail. Dccproc does not support any thresholds of its own, because equivalent effects can be achieved with regular expressions and you can apply dccproc several times using different DCC servers and then score mail based what all of the DCC servers say.

Error messages are sent to stderr as well as the system log. Connect stderr and stdout to the same file to see errors in context, but direct stderr to /dev/null to keep DCC error messages out of the mail. The -i option can also be used to separate the error messages.

Dccproc sends reports of checksums related to mail received by DCC clients and queries about the total number of reports of particular checksums. A DCC server receives no mail, address, headers, or other information, but only cryptographically secure checksums of such information. A DCC server cannot determine the text or other information that corresponds to the checksums it receives. It only acts as a clearinghouse of counts of checksums computed by clients.

For the sake of privacy for even the checksums of private mail, the checksums of senders of purely internal mail or other mail that is known to not be unsolicited bulk can be listed in a whitelist to not be reported to the DCC server.

When sendmail(8) is used, dccm(8) is a better DCC interface. Dccifd(8) is more efficient than dccproc because it is a daemon, but that has costs in complexity. See dccsight(8) for a way to use previously computed checksums.


The following options are available:
displays the version of the DCC procmail(1) interface.
enables debugging output from the DCC client library. Additional -d options increase the number of messages. One causes error messages to be sent to STDERR as well as the system log.
adds to existing X-DCC headers (if any) of the brand of the current server instead of replacing existing headers.
only queries the DCC server about the checksums of messages instead of reporting and then querying. This is useful when dccproc is used to filter mail that has already been reported to a DCC server by another DCC client such as dccm(8). No single mail message should be reported to a DCC server more than once per recipient.

It is better to use MXDCC lines in the global whiteclnt file for your MX servers

outputs only the X-DCC header and the checksums for the message.
outputs only the X-DCC header.
adds lines to the start of the log file turned on with -l and -c describing what might have been the envelope of the message. The information for the inferred envelope comes from arguments including -a and headers in the message when -R is used. No lines are generated for which no information is available, such as the envelope recipient.
says the first Received lines have the standard "helo (name [address])..." format and the address is that of the SMTP client that would otherwise be provided with -a The -a option should be used if the local SMTP server adds a Received line with some other format or does not add a Received line. Received headers specifying IP addresses marked MX or MXDCC in the -w whiteclnt file are skipped.
-h homedir
overrides the default DCC home directory, which is often /var/lib/dcc.
-m map
specifies a name or path of the memory mapped parameter file instead of the default map in the DCC home directory. It should be created with the new map operation of the cdcc(8) command.
-w whiteclnt
specifies an optional file containing SMTP client IP addresses and SMTP headers of mail that do not need X-DCC headers and whose checksums should not be reported to the DCC server. It can also contain checksums of spam. If the pathname is not absolute, it is relative to the DCC home directory. Thus, individual users with private whitelists usually specify them with absolute paths. Common whitelists shared by users must be in the DCC home directory or one of its subdirectories and owned by the set-UID user of dccproc It is useful to include a common or system-wide whitelist in private lists.

The format of the dccproc whiteclnt file is the same as the whitelist file required by dbclean(8) and dccm(8). Unlike dccm, the dccproc whiteclnt file is optional. When -w is not used, settings equivalent to these are used:

option log-normal
option dcc-on
option DCC-reps-on
option DNSBL-on

When -w is used, the defaults mentioned in dcc(8) are used. Those defaults differ and turn off DCC Reputations, and DNS blacklist (DNSBL )checking.

Because the contents of the whiteclnt file are used frequently, a companion file is automatically created and maintained. It has the same pathname but with an added suffix of .dccw It contains a memory mapped hash table of the main file.

A local whitelist entry ("OK) or two or more semi-whitelistings ("OK2") for one of the message's checksums prevents all of the message's checksums from being reported to the DCC server and the addition of a X-DCC header line by dccproc Because it is run by or on behalf of a single user, dccproc ignores env_To entries in the whiteclnt file. Users who don't want to use dccproc shouldn't.

-T tmpdir
changes the default directory for temporary files from the system default. The system default is often /tmp
-a IP-address
specifies the IP address (not the host name) of the immediately previous SMTP client. It is often not available. -a is ignored. -a The -a option should be used instead of -R if the local SMTP server adds a Received line with some other format or does not add a Received line.
-f env_from
specifies the RFC 821 envelope "Mail From" value with which the message arrived. It is often not available. If -f is not present, the contents of the first Return-Path: or UNIX style From_ header is used. The env_from string is often but need not be bracketed with "<>".
-t targets
specifies the number of addressees of the message if other than 1. The string many instead of a number asserts that there were too many addressees and that the message is unsolicited bulk email.
-x exitcode
specifies the code or status with which dccproc exits if the -c thresholds are reached or the -w whiteclnt file blacklists the message, unless the message is whitelisted.

The default value is EX_NOUSER. EX_NOUSER is 67 on many systems. Use 0 to always exit successfully.

-c type, [log-thold, ] rej-thold
sets logging and "spam" thresholds for checksum type Each logged message placed in a separate file in the directory specified with -l The checksum types are IP env_From From Message-ID substitute Received Body Fuz1 Fuz2 rep-total and rep The first six, IP through substitute have no effect except when a local DCC server configured with -K is used. The substitute thresholds apply to the first substitute heading encountered in the mail message. The string ALL sets thresholds for all types, but is unlikely to be useful except for setting logging thresholds. The string CMN specifies the commonly used checksums Body Fuz1 and Fuz2 Rej-thold and log-thold must be numbers, the string NEVER or the string MANY indicating millions of targets. Counts from the DCC server as large as the threshold for any single type are taken as sufficient evidence that the message should be logged or rejected.

Log-thold is the threshold at which messages are logged. It can be handy to log messages at a lower threshold to find solicited bulk mail sources such as mailing lists. Messages that reach at least one of their rejection thresholds or that have complicated combinations of white- and blacklisting are logged regardless of logging thresholds.

Rej-thold is the threshold at which messages are considered "bulk," and so should cause the X-DCC header line to contain the string "bulk" or "bulk rep" and dccproc to exit with the value set by -x

DCC reputation thresholds in the commercial version of the DCC are controlled by thresholds on checksum types rep and rep-total Messages from an IP address that the DCC database says has sent more than rep-total log-thold messages are logged. A DCC reputation is computed for messages received from IP addresses that have sent more than rep-total rej-thold messages. The DCC reputation of an IP address is the percentage of its messages that have been detected as bulk, or having at least 10 recipients. The defaults are equivalent to -c rep,never and -c rep-total,never,10

The checksums of locally white-listed messages are not checked with the DCC server and so only the number of targets of the current instance of a white-listed message are compared against the thresholds.

The default is -c ALL,NEVER so that nothing is discarded or logged. A common choice is -c CMN,25,50 to reject or discard mail with common bodies except as overridden by the whitelist of the DCC server and -g and -w

-g [not- ] type
indicates that white-listed, OK or OK2 counts from the DCC server for a type of checksum are to be believed. They should be ignored if prefixed with not- Type is one of the same set of strings as for -c Only IP env_From and From are likely choices. By default all three are honored, and hence the need for not-
-S hdr
adds to the list of substitute or locally chosen headers that are checked with the -w whiteclnt file and sent to the DCC server. The checksum of the last header of type hdr found in the message is checked. As many as 6 different substitute headers can be specified, but only the checksum of the first of the 6 will be sent to the DCC server.
-i infile
specifies an input file for the entire message instead of standard input. If not absolute, the pathname is interpreted relative to the directory in which dccproc was started.
-o outfile
specifies an output file for the entire message including headers instead of standard output. If not absolute, the pathname is interpreted relative to the directory in which dccproc was started.
-l logdir
specifies a directory for copies of messages whose checksum target counts exceed -c thresholds. The format of each file is affected by -E

If logdir starts with D? log files are put into subdirectories of the form logdir/JJJ where JJJ is the current julian day. H?logdir puts logs files into subdirectories of the form logdir/JJJ/HH where HH is the current hour. M?logdir puts log files into subdirectories of the form logdir/JJJ/HH/MM where MM is the current minute. See the FILES section below concerning the contents of the files. See also the option log-subdirectory-{day,hour,minute} lines in whiteclnt files described in dcc(8).

The directory is relative to the DCC home directory if it is not absolute

-B dnsbl-option
enables DNS blacklist checks of the SMTP client IP address, SMTP envelope Mail_From sender domain name, and of host names in URLs in the message body. Body URL blacklisting has too many false positives to use on abuse mailboxes. It is less effective than greylisting with dccm(8) or dccifd(8) but can be useful in situations where greylisting cannot be used.

Dnsbl-option is either of the forms set:option or domain [,IPaddr [,bltype ] ] Domain is a DNS blacklist domain such as example.com that will be searched. IPaddr is the string "any" or the IP address in the DNS blacklist that indicates that the mail message is spam. is assumed if IPaddr is absent. IPv6 addresses can be specified with the usual colon (:) notation. Names can be used instead of numeric addresses. The type of DNS blacklist is specified by bltype as name IPv4 or IPv6 Given an envelope sender domain name or a domain name in a URL of spam.domain.org and a blacklist of type name spam.domain.org.example.com will be tried. Blacklist types of IPv4 and IPv6 require that the domain name in a URL be resolved into an IPv4 or IPv6 address. The address is then written as a reversed string of decimal octets to check the DNS blacklist, as in,

More than one blacklist can be specified. They are searched in order. All searching is stopped at the first positive result.

Positive results are ignored after being logged unless an option DNSBL-on line appears in the global or per-user whiteclnt file.

-B set:debug=X
sets the DNS blacklist logging level
-B set:msg-secs=S
limits dccproc to S seconds total for checking all DNS blacklists. The default is 25.
-B set:URL-secs=S
limits dccproc to at most S seconds resolving and checking any single URL. The default is 11. Some spam contains dozens of URLs and that some "spamvertised" URLs contain host names that need minutes to resolve. Busy mail systems cannot afford to spend minutes checking each incoming mail message. In order to use typical single-threaded DNS resolver libraries, dccm(8) and dccifd(8) use fleets of helper processes.
-B set:no-envelope
says that SMTP client IP addresses and sender Mail_From domain names should not be checked in the following blacklists. set:envelope restores the default for subsequently named blacklists.
-B set:no-body
says that URLs in the message body should not be checked in the in the following blacklists. set:body restores the default for later blacklists.
-B set:no-MX
says MX servers of sender Mail_From domain names and host names in URLs should not be checked in the following blacklists. set:MX restores the default.
-B set:no-NS
says NS servers of sender Mail_From domain names and host names in URLs should not be checked in the following blacklists. set:NS restores the default.
-L ltype,facility.level
specifies how messages should be logged. Ltype must be error or info to indicate which of the two types of messages are being controlled. Level must be a syslog(3) level among EMERG ALERT CRIT ERR WARNING NOTICE INFO and DEBUG Facility must be among AUTH AUTHPRIV CRON DAEMON FTP KERN LPR MAIL NEWS USER UUCP and LOCAL0 through LOCAL7 The default is equivalent to
Something like this turns off the log messages:
-L notice,MAIL.debug -L error,MAIL.DEBUG

dccproc exits with 0 on success and with the -x value if the -c thresholds are reached or the -w whiteclnt file blacklists the message. If at all possible, the input mail message is output to standard output or the -o outfile despite errors. If possible, error messages are put into the system log instead of being mixed with the output mail message. The exit status is zero for errors so that the mail message will not be rejected.

If dccproc is run more than 500 times in fewer than 5000 seconds, dccproc tries to start Dccifd(8). The attempt is made at most once per hour. Dccifd is significantly more efficient than . With luck, mechanisms such as SpamAssassin will notice when dccifd is running and switch to dccifd.


DCC home directory in which other files are found.
memory mapped file in the DCC home directory of information concerning DCC servers.
contains the client whitelist in the format described in dcc(8).
is a memory mapped hash table corresponding to the whiteclnt file.
contains temporary files created and deleted as dccproc processes the message.
is an optional directory specified with -l and containing marked mail. Each file in the directory contains one message, at least one of whose checksums reached one of its -c thresholds. The entire body of the SMTP message including its header is followed by the checksums for the message.


The following procmailrc(5) rule adds an X-DCC header to passing mail
 :0 f
 | /usr/bin/dccproc -ERw whiteclnt

This procmailrc(5) recipe rejects mail with total counts of 10 or larger for the commonly used checksums:

 :0 fW
 | /usr/bin/dccproc -ERw whiteclnt -ccmn,10
 :0 e


cdcc(8), dcc(8), dbclean(8), dccd(8), dblist(8), dccifd(8), dccm(8), dccsight(8), mail(1), procmail(1).


Implementation of dccproc was started at Rhyolite Software in 2000. This describes version 1.3.48.


dccproc uses -c where dccm(8) uses -t